Hybrid data management strategy for enterprise AI success

The strength and agility of an organization's data strategy determine the success of an AI initiative just as much as advanced models and advanced algorithms do. AI agents and generative AI systems now drive business operations, which makes data accessibility and trustworthiness across on-premises, private and public cloud environments essential for business success.

Research from Enterprise Strategy Group, now part of Omdia, found that a mere 48% of organizations have enough trust in their data to confidently apply it to AI systems. This lack of trust underscores a deeper issue: Data remains segregated, fragmented, inconsistent, and often lacks proper governance. When organizations look to tap into this data for AI projects, they find it's not ready to serve as a foundation. The success of every AI use case depends on building a base of data. For example, if an organization is building a customer-facing knowledge base, inaccurate data will result in untrusted AI responses and damage to customer satisfaction. It's also critical to consolidate relevant data across cloud and on-premises environments. Two-thirds of data still resides on-premises for organizations, and AI might rely on both structured and unstructured data.

A hybrid data management strategy allows organizations to access all their data sources regardless of location and maintain seamless access and governance across data stored in legacy on-premises systems and secure private and flexible public cloud platforms. This makes it especially important to train large language models and deploy AI agents that depend on diverse real-time data sources.

Cloudera's True Hybrid vision exemplifies this strategy. The system delivers a unified data fabric across all environments, allowing data to move securely and without interruption. Organizations achieve smoother operations across multiple environments through universal security policies, centralized metadata management and centralized governance, along with the required data warehouse, data lake and data management capabilities. Cloudera AI also provides access to test and develop AI applications with the leading models and technology, such as Nvidia microprocessors.

The need for this flexibility is clear. Enterprise Strategy Group research found that 66% of organizations currently maintain or wish to pursue a hybrid data management approach. Why? Because AI isn't a one-size-fits-all model. Real-time fraud detection represents one example of use cases that require immediate access to on-premises data. Generative AI applications for customer engagement gain advantages from scalable cloud infrastructure along with access to extensive data pools.

AI's future is inherently hybrid. As AI integration expands across multiple organizational departments like HR, marketing, finance and logistics, organizations face increasing demands to maintain consistent and reliable data throughout all their operational environments. Organizations with on-premises data infrastructure need to adopt a hybrid strategy because it serves as a vital business requirement rather than merely a technological option. Rather than bringing data to AI, a strategy of bringing AI to your data can be far more efficient and cost-effective. Data movement can be expensive in hard costs, along with added complexity, governance and security challenges.

Oracle is also addressing some of the significant hybrid data challenges with the Oracle 23ai database and Autonomous 23ai managed offering. This brings hybrid database capabilities across environments and delivers high performance at scale when running on the Oracle Exadata platform. With Oracle's hyperscale partners like Google Cloud and Microsoft Azure, Oracle 23ai is more widely available. With so many AI use cases requiring structured and unstructured data, a unified database platform addresses both database management requirements with unified security.

Hybrid data management is not new for many organizations, but with AI, many believe that it lives in the cloud. Top technology vendors understand enterprise AI might have started in the cloud, but it increasingly needs to be where the data lives.

Enterprises are shifting toward hybrid data and AI management strategies to control costs, security and governance while having scalable access to all the AI resources they need. By adopting hybrid data platforms such as Cloudera and Oracle, organizations gain the agility and control needed to establish reliable data foundations to power AI projects. This approach supports the creation of unique AI applications, powered by contextual enterprise data that stand out as innovative while maintaining sustainability and scalability for successful outcomes.

Stephen Catanzano is a senior analyst at Enterprise Strategy Group, now part of Omdia, where he covers data management and analytics.

Enterprise Strategy Group is part of Omdia. Its analysts have business relationships with technology vendors.